Characterization of a scene and objects within the scene has traditionally involved user interaction during calibration. For example, the user may be required to place calibration targets or make and input manual measurements of the physical environment. Using the calibration targets, manual measurements, or both, a physical arrangement of the scene may be reconstructed. However, this calibration is time consuming, difficult, and constrained to the position at which the scanning equipment resided during the calibration.